Ask a plant manager what their worst unplanned downtime event cost, and you'll usually get a fairly precise number. Ask which part caused it, and the answer is often a little embarrassing. A servo encoder. A fieldbus communication module. A proximity sensor with an unusual connector. Rarely is it the motor, the gearbox, or anything that shows up on a capital budget.
That gap — between the cost of the failure and the cost of the part — is where most procurement decisions around industrial MRO suppliers go wrong. Maintenance defines the problem in technical terms. Purchasing evaluates it in unit-price terms. Both are doing their jobs correctly, and the plant still loses three shifts.
The purpose here isn't to argue for more inventory or more approved vendors. It's to look at how experienced teams actually decide who they buy from, and what questions separate a supplier who can help during a Tuesday-night breakdown from one who can only quote a catalog number.
The vendor list you have is not the vendor list you need
Most plants have somewhere between 200 and 2,000 active MRO vendors. Almost all of that volume — bearings, fasteners, filters, lubricants, standard pneumatics — moves through a handful of large distributors, and that arrangement works fine. Those items are commoditized, stocked domestically, and available same-day.
The trouble is that the same vendor list gets used for a completely different class of problem: the discontinued drive, the legacy PLC I/O card, the OEM-specific hydraulic manifold, the imported gearbox from a European machine builder. These don't behave like MRO consumables. They behave like project procurement compressed into 48 hours.
When a buyer sends that request to a general distributor, the usual response is a long quoted lead time from the manufacturer, or nothing at all. Not because the distributor is incompetent, but because their model is built around stocked SKUs and standard replenishment. Finding a discontinued Siemens or Allen-Bradley module in surplus channels, verifying firmware revision compatibility, and shipping it from overseas in three days is a different business entirely.
Plants that handle breakdowns well have usually recognized this and built a second tier of sourcing relationships for hard-to-find and obsolete components — separate from their primary MRO distribution agreements, and qualified in advance rather than found through a frantic Google search at 11 p.m.
What actually distinguishes a capable supplier
Price comparison is easy and mostly uninformative at this level. A few things are harder to evaluate and far more predictive.
Do they tell you the real lead time, or the catalog lead time? There's a meaningful difference between "manufacturer quotes 22 weeks" and "manufacturer quotes 22 weeks, but we've seen three of these clear in about seven from the regional stock, and here's what that depends on." The second answer requires someone who tracks actual fulfillment rather than reading a screen. It's also the answer that lets you decide whether to plan around a workaround or wait.
Can they trace where the part came from? For surplus, refurbished, and gray-market components, provenance is the whole question. A serious supplier can tell you whether a unit is factory-sealed, new-old-stock, tested-and-refurbished, or pulled from a decommissioned line — and can supply test reports, date codes, and a stated warranty that matches the condition. If the answer is vague, assume the worst case and price the risk accordingly.
Do they push back on your part number? Part-number verification sounds trivial until you've received a technically correct module that won't communicate with the rest of the rack because of a hardware revision difference. Suppliers who ask what it's connecting to, what firmware the CPU is running, or whether the machine has been retrofitted are doing engineering work on your behalf. The ones who simply confirm your number and take the order are transferring that risk to you.
What happens when it fails on arrival? Ask about the process, not the policy. Return authorization on a $60,000 drive that arrived DOA in the middle of a shutdown is a different conversation than a warranty clause on paper.
Lead time is information, not a number
One of the more useful habits a maintenance and procurement team can build together is treating lead time as a risk attribute of the equipment, not a field on a purchase order.
Take any critical asset and ask: if this component fails today, how long until production restarts? Not how long to get a quote — how long to get the line running. That number includes sourcing, customs if it's imported, shipping mode, any bench testing, and the labor window to install it.
Do this across a line and a pattern emerges quickly. Most components fall into a 24-to-72-hour bucket and need no special handling. A small number — usually under thirty parts on a typical line — carry exposure measured in weeks or months. Those are the ones that justify stock, a qualified alternate source, an approved substitute, or an engineering plan.
This exercise is worth more than a generic ABC inventory classification, because it's driven by downtime consequence rather than annual spend. A $200 safety relay with a 14-week replenishment window and no drop-in equivalent deserves more attention than a $9,000 motor available from four distributors in Ohio.
OEM, aftermarket, or refurbished — decide before the failure
The instinct to default to OEM is understandable and often correct, particularly where warranty coverage on the machine is still live, where the component sits in a safety-rated circuit, or where regulated documentation is required. But applied universally, it creates its own exposure: single-source dependency on a manufacturer who may have discontinued the product line or moved production overseas.
The realistic question isn't OEM versus aftermarket in the abstract. It's whether a specific alternative meets the requirements of a specific application.
For a standard three-phase motor driving a conveyor, an equivalent-frame replacement from a qualified manufacturer is usually a non-event. For a proprietary servo amplifier tuned to a machine builder's control architecture, an "equivalent" often means a week of commissioning you didn't budget for. Refurbished units occupy a sensible middle ground for discontinued electronics — frequently they're the only option that exists — provided the supplier can document what was tested and what was replaced.
Where teams get into trouble is making this determination under pressure. The engineering evaluation of an alternate source takes a couple of hours when the line is running and a couple of days of arguing when it isn't. Reviewing substitutes for your top-exposure components during a planned shutdown, and recording the approved alternates in the CMMS alongside the primary part number, removes that decision from the critical path entirely.
Obsolescence needs a procurement answer before it needs an engineering one
Control systems installed in the early 2000s are now well past manufacturer support in many plants. Full migration is the right long-term answer and is also a capital project with a queue, a budget cycle, and a commissioning window that competes with production.
In the meantime, somebody has to keep the existing system running. That's a sourcing problem. Securing a reasonable buffer of the remaining critical modules through surplus and secondary channels doesn't solve obsolescence, but it buys the eighteen or twenty-four months an orderly migration actually takes — which is considerably cheaper than an emergency retrofit forced by a failure with no spare.
This is one of the areas where firms working across broader international procurement channels, including specialist sourcing operations such as KTB Europe, tend to have visibility that a domestic distributor doesn't, simply because legacy European automation stock often sits closer to the original machine builders. For US plants running imported equipment, that geography matters more than it should.
A qualification process you can run this quarter
None of this requires a new system. It requires about a day of coordinated effort:
- Have maintenance list the components that would stop a line and cannot be replaced within 72 hours. Cap the list at fifty parts.
- For each, record the current source, the actual last-observed lead time, and whether an approved alternate exists.
- Identify which items are discontinued or approaching end-of-support.
- Test two or three specialist suppliers on that list with a real quote request, and evaluate the response quality — not just the price.
- Document approved alternates where they exist, in the maintenance system where a technician will actually find them.
The output is a short, specific list of exposures and a small set of pre-qualified relationships for the situations your primary distribution agreements weren't designed to cover.
Working with capable MRO procurement partners doesn't eliminate downtime. It changes what happens in the first two hours after a failure — whether that time is spent locating a part or waiting for someone to call back. Over a year, that difference usually costs more than any of the unit-price savings the sourcing strategy was built to capture.
FAQ
1. How many suppliers should we qualify for critical spare parts?
For genuinely critical components, two qualified sources is the practical minimum — a primary and one alternate capable of fulfilling under pressure. Beyond three, administrative burden outweighs the benefit. The alternate doesn't need volume; it needs to be pre-qualified so no approval process stands between a failure and a purchase order.
2. Is refurbished automation equipment acceptable for production-critical applications?
Often yes, particularly for discontinued components where no new stock exists. The determining factors are what testing was performed, what was replaced, and what warranty is offered. Refurbished units generally aren't appropriate in safety-instrumented functions or where regulatory documentation requires new, traceable parts.
3. How do we evaluate industrial MRO suppliers without running a formal RFQ?
Send two or three real quote requests for difficult parts you already have in hand. Evaluate how specific they are about lead time, whether they verify compatibility details, and how they describe part condition and provenance. That reveals far more than a questionnaire.
4. Should we stock inexpensive parts just because they have long lead times?
Yes, when absence stops production. Stocking decisions should follow downtime consequence rather than unit cost. A $150 component with a 12-week replenishment window and no substitute is a stronger stocking candidate than an expensive item available locally within a day.
5. What's the most common mistake in emergency sourcing?
Ordering on part number alone without confirming hardware revision, firmware version, or configuration differences. The part arrives correct on paper and doesn't function in the installed system — costing another full lead-time cycle at exactly the moment there's no time left.
Comments
Post a Comment